Previous reports have examined several proposed markers of oxidative stress in a model system of CCl4 administered to rats. Our work on the CCl4 model has identified a lower GSH/GSSG ratio as an early marker of oxidative stress, and lipid peroxidation products of plasma and urine MDA and F2-isoprostanes were reported as the most reliable, sensitive and accurate markers for measurement of oxidative stress since they were both increased in a time-and dose-dependent pattern. In the present study, rodent inhalation exposure to ozone was chosen as the second model of oxidative stress to determine whether exposure to ozone causes oxidative lung damage that can be quantitated by the identification and measurement of the presence of various products of oxidation in blood, plasma or urine. Although the molecular mechanisms of ozone-induced lung cell injury are not fully understood, one hypothesis is that ozone reacts with cell membrane lipids, inducing lipid peroxidation, aldehyde production and generation of lipid ozonation products that are considered to be involved in the early stage of inflammatory response. The first objective of this study was to determine whether acutely exposing rats to ozone would result in the loss of antioxidants from plasma and bronchoalveolar lavage fluid (BALF). Additional goals were to compare analyses of the same antioxidant concentration between different laboratories, to investigate which methods have the sensitivity to detect decreased levels of antioxidants, and to identify a reliable measure of oxidative stress in ozone-exposed rats. Male Fisher rats were exposed to either 2.0 or 5.0 ppm ozone inhalation for 2 h. Blood plasma and BALF samples were collected 2, 7, and 16 h after the exposure. It was found that ascorbic acid in plasma collected from rats after the higher dose of ozone was lower at 2 h, but not later. BALF concentrations of ascorbic acid were decreased at both 2 and 7 h post-exposure. Tocopherols (, , ), 5-nitro--tocopherol, tocol, glutathione (GSH/GSSG), and cysteine (Cys/CySS) were not decreased, regardless of the dose or post-exposure time point used for sample collection. Uric acid was significantly increased by the low dose at 2 h and the high dose at the 7 h point, probably because of the accumulation of blood plasma in the lung from ozone-increased alveolar capillary permeability. We conclude that measurements of antioxidants in plasma are not sensitive biomarkers for oxidative damage induced by ozone and are not a useful choice for the assessment of oxidative damage by ozone in vivo. The second objective of the ozone exposure study was to determine whether oxidation products of lipids, proteins and DNA could be identified as markers of oxidative stress. The time- and dose-dependent effects of ozone exposure on rat plasma lipid hydroperoxides, malondialdehyde, F2-isoprostanes, protein carbonyls, methionine oxidation, tyrosine and phenylalanine oxidation products, as well as urinary malondialdehyde and F2-isoprostanes, were investigated again with various techniques. The criterion used to recognize a marker in the model of ozone exposure was whether or not a significant effect could be identified and measured in a biological fluid seen at both doses at more than one time point. No statistically significant differences between the experimental and control groups at either ozone dose or time point studied could be identified in this study. Tissue samples were not included. Despite all the work accomplished in the BOSS study of ozone, no available product of oxidation in biological fluid has yet met the required criterion of being a biomarker. The current negative findings as a consequence of ozone exposure are of great importance, because they document that in complex systems, as in the present in vivo experiment, the assays used may not provide meaningful data of ozone oxidation, especially in human studies. The major conclusion of these studies is that the detection of oxidative stress markers in animal models is exceedingly difficult even in models near the LD50 level. These publications, which are highly cited, have undoubtedly saved other investigators both time and money, especially in epidemiological studies. The biomarker 8-iso-prostaglandin F2&#945; (8-iso-PGF2&#945;) is regarded as the gold standard for detection of excessive chemical lipid peroxidation in humans. We found that increase in 8-iso-PGF2&#945; do not necessarily reflect in oxidative stress; therefore, past studies using 8-iso-PGF2&#945; as a marker of oxidative stress may have been misinterpreted. The 8-iso-PGF2&#945;/PGF2&#945; ratio can be used to distinguish biomarker synthesis pathways and thus confirm the potential change in oxidative stress in the myriad of disease and chemical exposures known to induce 8-iso-PGF2&#945;.